Alexandra,
The puzzling warning you get is somewhat explained. The scale of at least
one of the coefficients was incompatible with running the hausman test.
Here, the coefficient of AGESQ is possibly too small compared to the other
coefficients. Some rescaling of AGESQ (and AGE) is probably required (for
instance multiplying these by 10) before rerunning the test.
Nonetheless, the null hypothesis of the hausman test - that there is no
systematic difference in the fixed effect and random effects coefficient
estimates is rejected here, which (might) suggest that the fixed effects
model (or conditional logit) is preferable.
However, regardless of the test findings, choice of fixed- or random effects
model is never straightforward, see for instance Gujarati (2003) p.650-651,
Green (2000) p.576-577 or the Stata manuals.
Best,
Johan Hellstrom
-------
Greene, W. H. (2000). Econometric analysis (fourth edition). Upper Saddle
River : NJ: Prentice-Hall.
Gujarati, D. N. (2003). Basic Econometrics (fourth edition). New York: NY.:
McGraw-Hill.
-------
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
